Loading profile
Loading sourced facts.
01 Finding sources
02 Reading evidence
03 Filing the card
Loading profile
01 Finding sources
02 Reading evidence
03 Filing the card
Datacurve supplies high-quality coding datasets to foundation model labs for post-training AI models.
Raised $18M across disclosed rounds.[1][2][4][6][9]
Series A closed October 2025, led by Chemistry.[1][2][3][4][5][6]
post-money valuation is not publicly disclosed.
Named investors include Chemistry, Y Combinator, Afore Capital, Homebrew, Balaji Srinivasan, Pioneer Fund, SignalFire, Wing Venture Capital.[1][3][7][9]
Serena Ge, Co-Founder, CEO[1][10][11]
Charley Lee, Co-Founder[1][10][11]
Datacurve · Jul 2026[9]
Serena Ge · Jul 2026[10]
Charley Lee · Jul 2026[11]
Peter J. · Jul 2026[12]
Jul 2026 · uwaterloo.ca
reported
Jul 2026 · homebrew.co
reported
Datacurve supplies high-quality coding datasets to foundation model labs for post-training AI models.
Ask for one referenceable production customer.
Not a recommendation. The first thing this ledger cannot answer.
Scale AI · scale.com · Same buyer (foundation model labs) and same budget (training data procurement), but Scale AI is a generalist labeling outsourcer while Datacurve focuses specifically on expert coding data via a bounty marketplace.[1][2]
Mercor · mercor.com · Same buyer and budget; Mercor is reportedly courting a multi-billion valuation and competes for the same post-training data spend from AI labs.[1][2]
Surge AI · surgehq.ai · Same buyer and budget; Surge is a go-to for premium labeling but uses a contractor model rather than Datacurve's bounty-based, gamified approach.[1][2]
The customer proof is company-sourced only. No independent source in this ledger confirms it.[9]
Ask for one referenceable production customer.
Not a recommendation. The first thing this ledger cannot answer.
Filled in the side panel for invited readers. This card carries sourced facts only.
Jul 2026 · vctavern.com
reported
Jul 2026 · tbpndigest.com
reported